Presentation (CTI)

OBJECTIVES: Artificial intelligence (AI) has the potential to support opportunistic detection of pulmonary malignancies on routinely acquired chest X-rays performed for indications unrelated to lung cancer. This study aimed to estimate the clinical and economic impact of implementing AI-assisted chest X-ray interpretation using real-world data (RWD) from a pilot deployment in routine clinical practice.
METHODS: A decision-analytic model was developed to compare AI-supported early detection with delayed diagnosis of pulmonary malignancies. Model inputs were informed by RWD from a pilot implementation of an AI-based chest X-ray analysis system and by relative survival data from the Czech National Cancer Registry (SVOD). Survival estimates by disease stage were used to model outcomes for patients diagnosed following AI-supported detection versus delayed clinical diagnosis. Relative survival was extrapolated where required and adjusted for age and sex distribution. Clinical benefits were expressed as life-years gained (LYG) and quality-adjusted life-years gained (QALY). Economic outcomes included diagnostic costs per additional detected malignancy and the incremental cost-effectiveness ratio (ICER). Costs included AI-assisted diagnostic procedures and treatment costs associated with patients whose deaths were avoided through earlier diagnosis. Future health outcomes were discounted according to standard health economic practice in Czechia (3%).
RESULTS: In the base-case analysis AI-supported detection of pulmonary malignancies on routine chest X-rays generated 1.70 discounted incremental life-years gained (LYG) per patient, representing a 57.2% increase compared with delayed diagnosis. The ICER was lower than CZK 1.2 million per QALY, which is the usual threshold for reimbursement approval. Results were sensitive to assumptions regarding diagnostic delay and stage-specific survival but remained robust across plausible parameter ranges.
CONCLUSIONS: RWD-informed modelling suggests that AI-supported interpretation of routine chest X-rays may provide meaningful clinical benefits through earlier detection of pulmonary malignancies while representing an economically attractive diagnostic strategy. These findings support further evaluation of AI-assisted opportunistic cancer detection in routine clinical practice.